Friday, 2020-09-04

*** k_mouza has joined #openstack-ironic00:09
*** k_mouza has quit IRC00:14
*** jamesden_ has quit IRC00:23
*** jamesdenton has joined #openstack-ironic00:25
*** chihhsin has joined #openstack-ironic00:35
*** jamesden_ has joined #openstack-ironic00:46
*** jamesdenton has quit IRC00:47
*** tkajinam has quit IRC00:59
*** tkajinam has joined #openstack-ironic00:59
*** gyee has quit IRC01:32
*** yolanda has quit IRC02:07
*** rcernin has quit IRC02:09
*** ijw has quit IRC02:17
*** rcernin has joined #openstack-ironic02:18
*** ijw has joined #openstack-ironic02:18
*** rcernin has quit IRC02:18
*** rcernin has joined #openstack-ironic02:19
*** cdearborn has quit IRC02:58
*** Qianbiao has joined #openstack-ironic03:02
*** zzzeek has quit IRC03:17
*** zzzeek has joined #openstack-ironic03:18
jandersdoes anyone happen to have working examples of enrolling redfish kit into Ironic? https://docs.openstack.org/ironic/latest/admin/drivers/redfish.html doesn't seem to work: http://paste.openstack.org/show/797445/ and doing it through bifrost playbooks has it's own challenges if the OOB controller is running self-signed SSL...03:23
jandersto elaborate bit more on the challenges of the playbook approach: enroll playbook doesn't like 'redfish_verify_ca': False (refuses to take the inventory with this included) and I suspect that retro-fitting this option after the fact triggers this issue:  http://paste.openstack.org/show/797443/03:26
jandersit's probably something simple but I'm pretty stuck so any pointers would be greatly appreciated :)03:26
openstackgerritSteve Baker proposed openstack/metalsmith master: metalsmith_deployment role switch to metalsmith_instances  https://review.opendev.org/73788803:34
*** rloo has quit IRC03:42
openstackgerritKaifeng Wang proposed openstack/ironic master: Add missing log for clean failed  https://review.opendev.org/74986103:43
*** JamesBenson has quit IRC03:48
openstackgerritChristopher Dearborn proposed openstack/sushy master: Allow monitoring progress of a firmware update  https://review.opendev.org/74595004:04
openstackgerritChristopher Dearborn proposed openstack/ironic master: WIP: Redfish driver firmware update  https://review.opendev.org/74961904:04
*** Nisha_Agarwal has joined #openstack-ironic04:39
*** hjensas has joined #openstack-ironic05:32
openstackgerritparesh sao proposed openstack/ironic master: OOB one button secure erase for iLO5 based HPE Proliant servers.  https://review.opendev.org/74386805:32
arne_wiebalckGood morning, ironic!05:45
arne_wiebalckjanders: for redfish, you mean how to enroll the nodes or what to configure on the controllers?05:46
arne_wiebalckjanders: I have done this recently with a first node :)05:46
arne_wiebalckjanders: All manually for now, though. But I (hopefully) have some notes on what I did.05:47
*** tzumainn has quit IRC05:54
Qianbiaojanders maybe it's due to cmd client version?06:22
Qianbiaoit seems that error is thrown by openstack client, but not conductor06:23
Qianbiaomorning, arne_wiebalck06:23
arne_wiebalckHey Qianbiao o/06:28
arne_wiebalckjanders: sorry, had missed that you pasted the command and the error06:28
QianbiaoHey o/06:28
*** belmoreira has joined #openstack-ironic06:29
arne_wiebalckjanders: at a first glance, that looks good ... how about omitting the driver_info completely (and try to add it later)?06:29
Qianbiaoarne_wiebalck janders i thinks that is openstack client bug.06:31
arne_wiebalckQianbiao: yeah, looks like it06:31
Qianbiaoyou may add some log on your local /opt/stack/bifrost/lib64/python3.6/site-packages/openstack/config/loader.py file06:31
*** Nisha_Agarwal has quit IRC06:36
*** dsneddon has quit IRC06:38
Qianbiaoarne_wiebalck are u familiar with ironic bm node deploy with nova? Not sure why it did not auto switch network when deploying.06:44
Qianbiaowhen clean bm, the network is correct, dhcp happens on provisiong network. But when deploy it bind bm to target network(tenant network), not provisiong network.06:45
arne_wiebalckQianbiao: yes, we deploy with nova ... however: we use a single network (and the whole network part in nova is replaced with a big noop), so I don't think I am able to help :-/06:55
arne_wiebalckQianbiao: and on top of this we do not use Ironic's pxe infra but have a downstream patch to connect to our local infra06:55
arne_wiebalckQianbiao: so, network-wise we are non-standard06:56
Qianbiaoarne_wiebalck aha, ok06:56
*** jtomasek has joined #openstack-ironic06:58
iurygregorygood morning Ironic, Happy Friday!07:06
openstackgerritlikui proposed openstack/ironic-python-agent master: Remove six  https://review.opendev.org/74987707:06
Qianbiaohi iurygregory07:22
*** tosky has joined #openstack-ironic07:25
openstackgerritlikui proposed openstack/ironic-python-agent master: Remove six/babel/oslo.i18n from lower-constraints.txt  https://review.opendev.org/74987707:25
*** mgoddard has joined #openstack-ironic07:38
*** dougsz has joined #openstack-ironic07:39
*** rcernin_ has joined #openstack-ironic07:42
*** rcernin has quit IRC07:44
*** zzzeek has quit IRC07:45
*** zzzeek has joined #openstack-ironic07:47
rpittaugood morning ironic! o/07:48
iurygregoryhey rpittau o/07:51
rpittauhey iurygregory :)07:51
iurygregoryrpittau, the odata.id in redifh will always be "/redfish/v1/....." ?07:55
rpittauiurygregory: normally, yes07:55
iurygregorytks /me blames airship lib to access redfish now07:55
*** rcernin_ has quit IRC07:56
openstackgerritArun S A G proposed openstack/ironic-specs master: Add support for anaconda deploy interface  https://review.opendev.org/74850307:59
*** lucasagomes has joined #openstack-ironic08:06
openstackgerritAija Jaunteva proposed openstack/sushy master: Add BIOS update apply time and maintenance window  https://review.opendev.org/74923608:22
*** dtantsur|afk is now known as dtantsur08:24
dtantsurmorning ironic08:26
dtantsurjanders: weird, I'll take a look08:26
iurygregorygood morning dtantsur08:26
rpittauhey dtantsur :)08:26
*** derekh has joined #openstack-ironic08:27
dtantsurjanders: it's a bug in openstacksdk :(08:33
*** dsneddon has joined #openstack-ironic08:38
dtantsurjanders: try https://review.opendev.org/#/c/749903/08:43
patchbotpatch 749903 - openstacksdk - Fix a bogus error in config loader when using seve... - 1 patch set08:43
dtantsurfolks, let's pile on sushy reviews, we're in a pressing need of a final release08:50
dtantsurplease pay particular attention to features that are ready to go08:50
openstackgerritMerged openstack/ironic master: Enhance certificate verification for ilo harware type  https://review.opendev.org/74349008:51
dtantsurrealistically, it's probably only https://review.opendev.org/#/c/749236/08:54
patchbotpatch 749236 - sushy - Add BIOS update apply time and maintenance window - 2 patch sets08:54
iurygregoryyeah08:54
dtantsurI don't think I can grok the TaskMonitor patch so quickly08:54
iurygregoryI'm wondering about the https://review.opendev.org/#/c/748927/08:57
patchbotpatch 748927 - sushy - Make message parsing more resilient - 3 patch sets08:57
dtantsurseems contentious?08:57
dtantsurit's a bug fix, so can be backported when finished08:58
iurygregoryyeah08:58
openstackgerritAija Jaunteva proposed openstack/sushy master: Add BIOS update apply time and maintenance window  https://review.opendev.org/74923608:59
openstackgerritMerged openstack/bifrost stable/ussuri: Create our own firewalld zone and use it on real bare metal  https://review.opendev.org/74976009:02
openstackgerritMerged openstack/bifrost bugfix/8.3: Create our own firewalld zone and use it on real bare metal  https://review.opendev.org/74975909:08
iurygregorydtantsur, that problem with 'HTTPBasicAuth' object has no attribute 'get_auth_ref' your patch is merged?09:15
dtantsurnope09:15
iurygregorysad =(09:15
iurygregoryabout metalsmith I saw stevebaker said there is no problem if we can't merge his patch in the release, so we can move on and release metalsmith?09:17
dtantsuriurygregory: yeah. the releases have been proposed by the release team, so no actions needed from us09:20
*** dougsz has quit IRC09:22
iurygregorydtantsur, oh ok!09:23
*** aedc has joined #openstack-ironic09:23
openstackgerritDmitry Tantsur proposed openstack/ironic master: Do not assume that prepare_image is the last command to run  https://review.opendev.org/74991209:34
openstackgerritDmitry Tantsur proposed openstack/ironic stable/ussuri: Do not assume that prepare_image is the last command to run  https://review.opendev.org/74991309:38
*** dougsz has joined #openstack-ironic09:38
openstackgerritDmitry Tantsur proposed openstack/bifrost stable/ussuri: DNM test ussuri CI  https://review.opendev.org/74991409:39
openstackgerritYogesh Ramachandra proposed openstack/ironic master: [WIP] Fix idrac-wsman factory-reset getting updated bios setting values  https://review.opendev.org/74869609:39
openstackgerritDmitry Tantsur proposed openstack/ironic master: Add an option to require TLS for agent callback_url  https://review.opendev.org/74991610:11
*** Lucas_Gray has joined #openstack-ironic10:12
*** dougsz has quit IRC10:14
*** dougsz has joined #openstack-ironic10:30
*** Nisha_Agarwal has joined #openstack-ironic10:33
QianbiaoHello, to provision node through provisioning network, should the node's network interface must be "neutron"?10:47
QianbiaoCurrent, i am using flat, the node can be cleaned through provisioning network(clean network), but when deploy, it can not load dhcp ip.10:48
Nisha_AgarwalQianbiao, u can use ur own dhcp server also10:55
Nisha_Agarwaland dhcp_controller in the config file as "None"10:55
openstackgerritDmitry Tantsur proposed openstack/ironic master: Accept and use a TLS certificate from the agent  https://review.opendev.org/74713610:55
Nisha_AgarwalQianbiao, set the dhcp_controller in the config file as None10:56
dtantsurQianbiao: to have a separate provisioning network, you need to use the "neutron" network_interface10:57
dtantsurotherwise provisioning happens on the same flat tenant network10:57
Qianbiao<Nisha_Agarwal> i have thought that way. use a standalone dhcp server.10:59
QianbiaoBut when we have a big number nodes, it is not easy to maintence10:59
jandersarne_wiebalck Qianbiao dtantsur thanks heaps for looking into my issue10:59
Qianbiaodtantsur ok, i see, will try.10:59
jandersdtantsur wow you wrote a fix already - legend! will test it now11:00
Nisha_AgarwalQianbiao, ok. Then you can do as dtantsur said above ^^^^11:00
Nisha_Agarwaldtantsur, please see this patch https://review.opendev.org/#/c/746416/ when u have time11:01
patchbotpatch 746416 - ironic - Adds few of the security dashboard parameters to c... - 12 patch sets11:01
*** aedc has quit IRC11:04
*** derekh has quit IRC11:05
*** derekh has joined #openstack-ironic11:07
openstackgerritMerged openstack/ironic master: Trivial: fix minor typo on RFC number  https://review.opendev.org/74975411:18
*** zzzeek has quit IRC11:20
*** zzzeek has joined #openstack-ironic11:21
QianbiaoHi dtantsur, after set node network interface to neutron, I can not add node port manually, last error: http://paste.openstack.org/show/797459/11:22
jandersdtantsur with https://review.opendev.org/#/c/749903/ the node registers correctly (and I had one more issue with my syntax on top of hitting the bug)11:22
patchbotpatch 749903 - openstacksdk - Fix a bogus error in config loader when using seve... - 1 patch set11:22
Qianbiaodid I miss something, like should add extra info?11:22
jandershowever I'm hitting futher problems unfortunately, will paste the logs shortly (essentially the node gets stuck in 'enroll')11:23
jandersCLI output: http://paste.openstack.org/show/797458/11:23
janderslogs: http://paste.openstack.org/show/797457/11:23
jandersI suspect Ironic can't talk to the redfish interface of the node for some reason but I can't see why yet - I can use these creds and get into the iLO11:24
Qianbiao<janders>11:25
Qianbiaoto go on, you should update node status11:25
Qianbiaolike "openstack node manage node-name"11:25
jandersQianbiao I tried that, no effect unfortunately11:27
jandershttp://paste.openstack.org/show/797460/11:27
jandersmy first guess would be an issue with redfish details, but I can assemble a working curl call out of what I defined the node with...11:27
Qianbiaothe power status is None11:28
Qianbiaoit means the redfish driver did not connect to your server maybe?11:28
jandersyeah it really looks like it's not getting power info, but I dont see an indication of why11:28
jandersyeah I think you are right11:28
jandersdo you have any ideas how to get more details on why it's not connecting?11:29
arne_wiebalckjanders: did you interrupt the manage command?11:30
arne_wiebalckjanders: manage will not talk to the node I think11:30
arne_wiebalckjanders: hmm ... maybe it verifies sth when it does the state transition?11:30
jandersarne_wiebalck - no I don't recall interrupting it. It runs  cleanly (exit code 0) but takes no effect.11:31
arne_wiebalckjanders: in any case, if the transition fails, I'd expect to see sth in the logs11:31
arne_wiebalckjanders: there's a Ctrl-C in your paste, but that is maybe for the reverse search11:32
jandersarne_wiebalck yeah that would be the reverse search11:32
arne_wiebalckjanders: yes, there is some verification from enroll to manage11:33
jandersone thing it does react to is setting/unsetting maintenance11:33
jandersthat works11:33
arne_wiebalckjanders: this is just a flag in the DB I believe11:33
arne_wiebalckjanders: if the verification fails, the node will stay in enroll11:34
jandersarne_wiebalck I re-tried manage while tailing logs, got the same old error I was hitting before11:34
janders(pasting)11:34
jandershttp://paste.openstack.org/show/797463/11:35
jandersI've seen this while trying to 1) enroll the node via bifrost and 2) override the ssl_ignore after the fact with the CLI11:35
arne_wiebalckjanders: what about "openstack baremetal node validate <IRONIC NODE>" ?11:35
janderstrying now11:36
arne_wiebalckjanders: this should return sth if there is sth missing11:36
jandershttp://paste.openstack.org/show/797464/11:36
Qianbiao<janders> did u check whether redfish url requires /redfish/v111:37
Qianbiaothe log shows it returns a none json data.11:37
jandersQianbiao good question, I think it does:11:37
QianbiaoI think you may provide a wrong url.11:37
jandershttp://paste.openstack.org/show/797466/11:38
arne_wiebalckQianbiao: I don't think so, I have it configured like this as well.11:41
arne_wiebalckjanders: so, the verification is failing (from http://paste.openstack.org/show/797463/)11:42
arne_wiebalckjanders: this explains why the node does not move out of enroll11:42
jandersarne_wiebalck because of lack of defined ports?11:42
arne_wiebalckjanders: I don't think so11:43
arne_wiebalckjanders: did you curl to the redfish endpoint to see what the response is?11:43
jandersarne_wiebalck I tried a couple generic calls, it seems to work11:43
jandersarne_wiebalck should I try something specific?11:44
jandershmm11:45
jandersfrom this:  HTTP response for GET https://10.19.109.248/: status code: 20011:45
jandersI wonder if redfish_address should include the context11:46
arne_wiebalckjanders: curl -k -X GET http://10.19.109.248/redfish/v1/Systems11:46
jandersarne_wiebalck: http://paste.openstack.org/show/797468/11:47
Qianbiaojanders as i said11:48
Qianbiaochange your redfish url to xxx/redfish/v111:48
jandersQianbiao it took me a while to get it myself - good point, trying now11:49
jandersQianbiao arne_wiebalck there is some change11:53
jandershttp://paste.openstack.org/show/797469/11:54
jandersstill doesnt work but now we get sushy.exceptions.AccessError: HTTP GET https://10.19.109.248/redfish/v1/Systems/1 returned code 401. iLO.0.10.ExtendedInfo: See @Message.ExtendedInfo for more information.11:54
jandersI've seen this while trying to navigate around the redfish objects from the web browser11:54
janders{"error":{"code":"iLO.0.10.ExtendedInfo","message":"See @Message.ExtendedInfo for more information.","@Message.ExtendedInfo":[{"MessageId":"Base.1.0.NoValidSession"}]}}11:54
janders^^^ that's what it results in11:54
jandersfix: use curl and explicitly set auth11:55
jandersthen it works a treat11:55
jandershmmm now what to do with this11:55
*** Nisha_Agarwal has quit IRC11:56
Qianbiaojanders seems like sushy version issue.11:58
janderstoo old? too new?11:59
QianbiaoNot sure.12:00
rpittaujanders: can you double-check sushy version installed ?12:00
*** Nisha_Agarwal has joined #openstack-ironic12:01
jandersrpittau sushy (3.3.0) sushy-tools (0.12.0)12:01
jandersI have to say I am tempted to grab a Dell or a SuperMicro (preferrably Dell) and re-test12:02
rpittaujanders: you should, at least to have a second test to compare12:02
jandersI dont remember non-HPs kicking me out of redfish objects with 401 while playing around in browser12:02
rpittauversions look good12:02
janderswhich I think is what is happening to sushy/Ironic now12:02
Qianbiaojanders from your log it's pretty like a version incompatible issue.12:03
openstackgerritDmitry Tantsur proposed openstack/ironic-python-agent master: [EARLY WIP] Generate a TLS certificate and send it to ironic  https://review.opendev.org/74993012:08
*** zbitter is now known as zaneb12:10
arne_wiebalckjanders: on Train, I use python2-sushy-2.0.3-112:14
jandersallright, it's getting late here so time for me to call it a night12:25
jandershave a great weekend everyone!12:25
jandersthanks for all your help today, much appreciataed12:25
*** Nisha_Agarwal has quit IRC12:25
arne_wiebalckjanders: o/12:36
*** cdearborn has joined #openstack-ironic12:50
rpittauhey if someone has a moment, I did some cleaning on job names between ipa and ipa-builder https://review.opendev.org/746766 https://review.opendev.org/749723 https://review.opendev.org/74972513:00
patchbotpatch 746766 - ironic-python-agent - Prepare to move ipa-builder jobs - 3 patch sets13:00
patchbotpatch 749723 - ironic-python-agent-builder - Use converted ipa jobs - 2 patch sets13:00
patchbotpatch 749725 - ironic-python-agent - Remove unused jobs - 2 patch sets13:01
openstackgerritOpenStack Release Bot proposed openstack/metalsmith stable/victoria: Update .gitreview for stable/victoria  https://review.opendev.org/74993613:04
openstackgerritOpenStack Release Bot proposed openstack/metalsmith stable/victoria: Update TOX_CONSTRAINTS_FILE for stable/victoria  https://review.opendev.org/74993713:04
openstackgerritOpenStack Release Bot proposed openstack/metalsmith master: Update master for stable/victoria  https://review.opendev.org/74993813:04
openstackgerritOpenStack Release Bot proposed openstack/metalsmith master: Add Python3 wallaby unit tests  https://review.opendev.org/74993913:04
*** rloo has joined #openstack-ironic13:09
*** e0ne has joined #openstack-ironic13:32
e0necould anybody from core team take a look on a simple patch to ironic-ui please?13:33
TheJuliae0ne: sure13:40
TheJuliathe goal patch?13:40
e0neoops, didn't paste the link:(13:40
e0nehere is it https://review.opendev.org/#/c/744708/13:40
patchbotpatch 744708 - ironic-ui - [goal] Migrate testing to ubuntu focal - 16 patch sets13:40
e0neTheJulia: thank you13:40
TheJuliaapproved13:41
TheJulia:)13:41
dtantsurgood morning TheJulia13:41
e0negreat! we've one more "green" job on horizon gates now :)13:41
*** ironic-starter has joined #openstack-ironic13:42
ironic-starterI am a starter with stand-alone ironic and I have got to the point where I can bootstrap a server with CentOS8 Cloud Image13:43
ironic-starterI have a requirement where I need to set a password for teh cloud user centos13:43
ironic-starterhow do I do that13:43
ironic-starterI tried using configdrive but the password is never set in13:43
openstackgerritDmitry Tantsur proposed openstack/ironic-python-agent master: [WIP] Generate a TLS certificate and send it to ironic  https://review.opendev.org/74993013:44
*** Goneri has joined #openstack-ironic13:45
ironic-starterfollowed this https://docs.openstack.org/ironic/ussuri/install/configdrive.html and created a base64 enconded iso.gz13:45
dtantsurironic-starter: that's really about configdrive, maybe a problem with the syntax? if you have any way to look at the node's console (via its BMC, for example), you may see what cloud-init says13:45
ironic-starterdtantsur : does not say anything13:45
ironic-starterother than showing that the network interfaces are being configured and centos does not have any authorized ssh keys13:46
dtantsurI'd also recommend to provide a configdrive via the new syntax (also documented there) - smaller chance to make a mistake13:46
dtantsuri.e. where you provide a JSON13:46
ironic-starterdtantsur : you mean  " openstack baremetal node deploy $node_identifier \13:47
dtantsurironic-starter: a bit of a self-advertisement :) I've written a CLI tool/Python lib that simplifies some of these tasks: https://docs.openstack.org/metalsmith/latest/13:47
ironic-starterdtantsur : I have seen that project ... :)13:48
dtantsuryou can take a look at its code to see how to work with cloud-init user_data: https://opendev.org/openstack/metalsmith/src/branch/master/metalsmith/instance_config.py#L8813:48
ironic-starterdtantsur : Thanks a lot that helps13:49
ironic-starterI am specifically looking at setting the cloud user's password13:49
ironic-starterwould you have any specific inputs for me13:49
ironic-starterfor instance when I use this syntax13:50
ironic-starter#cloud-config13:50
ironic-starterwith the same cloud image without ironic just using kvm to boot that image it works13:50
ironic-starterI tried to provide the same syntax in user_data and now I think it is incorrect13:51
ironic-starterso looking forward for some guidance13:51
dtantsurI think you should be fine providing {"user_data": "#cloud-config\n<your conf here>"}13:51
dtantsurat least it works for metalsmith, maybe something goes wrong when building an ISO?13:51
openstackgerritMerged openstack/ironic-ui master: [goal] Migrate testing to ubuntu focal  https://review.opendev.org/74470813:53
QianbiaoHi dtantsur, may i take you a minute ~13:54
ironic-starteras per the document I create the following folder13:54
ironic-starter.13:54
ironic-starter2018-08-2713:54
ironic-starterand ran13:54
ironic-startermkisofs -R -V config-2 -o /opt/data/configdrive-8d030b3a-12a7-4f9b-9837-4fed2371b351.iso /root/configdrive/8d030b3a-12a7-4f9b-9837-4fed2371b35113:54
dtantsurironic-starter: do you *need* to go down this route?13:54
ironic-startergzip -c /opt/data/configdrive-8d030b3a-12a7-4f9b-9837-4fed2371b351.iso | base64 > /opt/data/configdrive-8d030b3a-12a7-4f9b-9837-4fed2371b351.iso.gz13:54
ironic-starterI dont13:55
dtantsurproviding a JSON leaves much less space for errors13:55
ironic-starterI am lost13:55
ironic-starterhence looking for a guidance13:55
ironic-starterIf there is a way that config-drive iso.gz can be generated on the fly that would be best13:55
dtantsurthat's literally what I'm talking about13:56
ironic-starterelse I would have to pass it on the cli13:56
ironic-startergot your point ... let me test and see13:56
dtantsursee the 2nd example in https://docs.openstack.org/ironic/ussuri/install/configdrive.html#when-used-standalone you can do the same with user_data13:56
ironic-starterdtantsur : let me test that and report back .. hope this works... I just didn't want to put in the user password on the command line ... but let me give it a shot and see13:58
dtantsurironic-starter: you're not putting a literal password there, if I recall correctly, but a hash of it13:58
ironic-startercloud-init file which I used with kvm boot had the password in plain-text ...13:59
ironic-starterbut when you say hash md5 or sha256 ?14:00
dtantsurironic-starter: https://cloudinit.readthedocs.io/en/latest/topics/examples.html explains it14:00
dtantsurlooks for 'passwd' on that page14:00
ironic-starterok14:01
dtantsurI don't see an ability to provide a plain text password, only a hash. the comments explain how to generate one.14:01
ironic-starterThanks :)14:02
ironic-starterlet me test now and see14:02
* dtantsur wonders if we should document it better14:03
*** uzumaki has joined #openstack-ironic14:03
openstackgerritMerged openstack/metalsmith master: Update master for stable/victoria  https://review.opendev.org/74993814:04
openstackgerritMerged openstack/metalsmith stable/victoria: Update .gitreview for stable/victoria  https://review.opendev.org/74993614:04
openstackgerritMerged openstack/metalsmith stable/victoria: Update TOX_CONSTRAINTS_FILE for stable/victoria  https://review.opendev.org/74993714:04
TheJuliaand it finally occurs to me Monday is a holiday here14:06
iurygregoryIn Brazil would be holiday also14:08
* iurygregory sad14:08
iurygregorygood morning TheJulia =)14:08
*** Qianbiao has quit IRC14:09
*** bfournie has joined #openstack-ironic14:09
*** Qianbiao has joined #openstack-ironic14:09
dtantsurlet's decide if we're ready to merge https://review.opendev.org/#/c/749236/ ?14:10
patchbotpatch 749236 - sushy - Add BIOS update apply time and maintenance window - 3 patch sets14:10
dtantsurwe cannot hold the sushy release forever14:11
*** zzzeek has quit IRC14:11
*** uzumaki has quit IRC14:12
iurygregorydtantsur, I think we can, I was expecting a green run from Dell CI but they are broken atm14:12
*** zzzeek has joined #openstack-ironic14:12
dtantsursigh14:12
iurygregoryneutron problems...14:12
dtantsurthis is a new feature, shouldn't be involved unless called explicitly14:12
iurygregoryyeah, I'm wondering if they tested downstream or something14:13
iurygregorythe code looks good14:13
*** jtomasek has quit IRC14:15
* TheJulia looks14:16
dtantsurunless I'm hallucinating, it is required for their redfish-based BIOS implementation14:17
TheJuliadtantsur: got a link?14:17
TheJuliaugh https://review.opendev.org/#/c/745950/ is a lot14:17
patchbotpatch 745950 - sushy - Allow monitoring progress of a firmware update - 15 patch sets14:17
dtantsuraha, https://review.opendev.org/#/c/749240/14:18
patchbotpatch 749240 - ironic - Add Redfish BIOS interface to idrac HW type - 2 patch sets14:18
TheJulialooking14:18
dtantsurthe sushy part in question is https://review.opendev.org/#/c/749236/14:18
patchbotpatch 749236 - sushy - Add BIOS update apply time and maintenance window - 3 patch sets14:18
TheJuliayup14:18
TheJuliait is required14:18
dtantsurI honestly gave up on hoping to squeeze the firmware update patch in the release14:18
dtantsurI don't have enough spoons for 700+ loc14:19
TheJuliaI do though14:19
TheJuliajust need to finish caffination14:19
TheJuliaanyone else up for 700 loc?14:19
dtantsurTheJulia: you want to try get it in Victoria or just?14:19
dtantsurthe last time I opened it I was worried how invasive it was14:20
TheJuliaIt is kind of clear to me that dell is trying to push a lot at the last minute... which is not great14:20
*** uzumaki has joined #openstack-ironic14:20
TheJuliaif it is invasive, then we really shouldn't at the last minute14:20
dtantsurnew arguments, renaming public methods..14:20
TheJuliaif it is "standard" pattern then it shouldn't be too bad14:20
TheJuliaoh jeeze14:20
dtantsurdunno, I'll rely on your judgement14:20
TheJuliayeah, no too late in the cycle for renaming methods if there is no compatability14:20
dtantsurjust keep in mind that the release deadline was yesterday :)14:21
TheJulia*shrugs*14:21
TheJuliaI think the real deadline is actually sunday, at least if I'm remembering what one of the notes said14:21
dtantsurI think it's always Thursday14:22
TheJuliaprocedural thursday and reality sunday ;)14:22
dtantsurunrelated: if you've ever dreamed of telling me I suck in security, this is your chance: https://review.opendev.org/#/c/749930/2/ironic_python_agent/tls_utils.py14:22
patchbotpatch 749930 - ironic-python-agent - [WIP] Generate a TLS certificate and send it to ir... - 2 patch sets14:22
TheJuliaheh14:22
iurygregoryI'm up for 700 loc14:26
TheJuliaUmmm... yeah I'm thinking this is a nope, but it otherwise is in good shape14:26
iurygregoryI can start now and finish after dinner probably XD14:27
TheJuliak14:27
iurygregorybecause in 1hr 30min I need to go see a flat14:27
TheJulia++14:27
Qianbiaohello, no means to interup you.14:30
Qianbiaoafter set node's network interface to neutron, I can not add node port manually, last error: http://paste.openstack.org/show/797459/14:31
Qianbiaoany experience to fix this14:31
iurygregorymaybe neutron logs will provide more reasons why it couldnt create the port14:34
Qianbiao<iurygregory> will check now14:35
*** ironic-starter has quit IRC14:37
openstackgerritDmitry Tantsur proposed openstack/ironic master: Accept and use a TLS certificate from the agent  https://review.opendev.org/74713614:37
*** priteau has joined #openstack-ironic14:43
TheJuliacdearborn: so https://review.opendev.org/#/c/745950 is missing a release note14:43
patchbotpatch 745950 - sushy - Allow monitoring progress of a firmware update - 15 patch sets14:43
TheJuliaany chance you can whip one up REALLY quick?14:43
openstackgerritDmitry Tantsur proposed openstack/ironic-python-agent-builder master: Install libssl-dev / openssl-devel when building IPA  https://review.opendev.org/74997514:43
cdearbornTheJulia: sure - will do as soon as i get off the phone with rpioso14:44
openstackgerritDmitry Tantsur proposed openstack/ironic-python-agent master: [WIP] Generate a TLS certificate and send it to ironic  https://review.opendev.org/74993014:46
Qianbiaoiurygregor no lucky..14:47
iurygregoryQianbiao, nothing in the logs?14:50
Qianbiaoyeah.14:50
QianbiaoI think maybe i miss some option which is required by "neutron" network port.14:50
Qianbiaomaybe lldp?14:50
Qianbiaoor physical network, i am not sure, did not find a document.14:51
TheJuliacdearborn: ack, thanks14:53
*** zzzeek has quit IRC14:56
*** zzzeek has joined #openstack-ironic14:59
TheJuliaerbarr: any luck on figuring out what is going on?14:59
erbarrTheJulia, nope I even used stable/ussuri for neutron and still same problem, i'm trying neutron-lib now15:00
openstackgerritJulia Kreger proposed openstack/ironic master: Detail iPXE + LACP troubleshooting information  https://review.opendev.org/74378715:01
TheJuliayeouch15:01
*** belmoreira has quit IRC15:03
Qianbiaowhat is the service name for neutron api in devstack?15:03
TheJuliaq-svc15:04
*** uzumaki has quit IRC15:05
*** uzumaki has joined #openstack-ironic15:06
Qianbiao<TheJulia> thanks.15:06
Qianbiaoiurygregory yes nothing.15:07
QianbiaoIs it possible that to provision in a l3 provisioning network with ovs?15:07
Qianbiaoor we need some special neutron driver15:08
QianbiaoI find it documented here. https://docs.openstack.org/ironic/rocky/admin/multitenancy.html15:08
Qianbiaoat last, it list supported drivers.15:08
*** uzumaki has quit IRC15:10
openstackgerritDmitry Tantsur proposed openstack/ironic-python-agent master: Fix and run the correct functional tests job  https://review.opendev.org/74998315:10
dtantsursigh ^^^15:10
dtantsurQianbiao: you can use an ML2 driver that supports OVS, but this is only for testing on VMs15:11
dtantsur(this includes networking-generic-switch and networking-baremetal)15:11
Qianbiaooh.....15:11
Qianbiaogreat news..15:12
arne_wiebalckGeneral bare metal ops question (outside Ironic): how do people manage the boot order changes in UEFI mode? Background: We usually assume hosts *always* boot off the network, which UEFI installations do not do, as they change the boot order at the end. We have a work around (to not run efibootmgr, but create a config, which will preserve the boot order), but we were wondering how this is handled in other15:12
arne_wiebalckdeployments ...15:12
Qianbiaodtantsur thanks, It means all i did now has no way to work.15:13
dtantsurgenerally, separate networks require a suitable ML2 driver to work with your switch15:14
dtantsurotherwise, ironic has no way to put nodes on different networks, and you have to use flat networking15:14
Qianbiaodtantsur i try flat networking.15:16
Qianbiaobut it does not support switch networks when deploy15:16
QianbiaoAm i right?15:16
openstackgerritChristopher Dearborn proposed openstack/sushy master: Allow monitoring progress of a firmware update  https://review.opendev.org/74595015:16
QianbiaoI can put provisioning network on a flat network.15:16
QianbiaoBut it did not provide dhcp when deploy. On provide dhcp fro clean.15:17
openstackgerritMerged openstack/sushy master: Add BIOS update apply time and maintenance window  https://review.opendev.org/74923615:17
rpittaubye everyone, see you in 4 weeks! o/15:18
*** rpittau is now known as rpittau|afk15:18
openstackgerritChristopher Dearborn proposed openstack/sushy master: Allow monitoring progress of a firmware update  https://review.opendev.org/74595015:20
cdearbornTheJulia: Added a reno to https://review.opendev.org/74595015:21
patchbotpatch 745950 - sushy - Allow monitoring progress of a firmware update - 17 patch sets15:21
cdearbornTheJulia: Accidentally pushed a change I didn't mean to, so had to undo that15:22
cdearbornTheJulia: should be good to go now15:22
cdearbornTheJulia: how would I go about requesting an FFE for firmware update support in the redfish driver?  I have a submitted patch with everything working: https://review.opendev.org/#/c/749619/.  Need to do test cases, docs, plus maybe a few more minor changes15:25
patchbotpatch 749619 - ironic - WIP: Redfish driver firmware update - 4 patch sets15:25
cdearbornTheJulia - ack!  don't know what i did - i just killed that patch.  will try to unravel15:31
*** jamesden_ is now known as jamesdenton15:31
cdearbornTheJulia - nevermind - patch is fine - need more coffee15:34
dtantsurI don't think we're in feature freeze yet?15:34
openstackgerritMerged openstack/ironic master: Add missing log for clean failed  https://review.opendev.org/74986115:40
openstackgerritMerged openstack/ironic master: Update deploy steps documentation  https://review.opendev.org/74972915:40
openstackgerritMerged openstack/ironic-lib master: Add Python3 wallaby unit tests  https://review.opendev.org/74978115:40
openstackgerritMerged openstack/ironic-lib stable/victoria: Update .gitreview for stable/victoria  https://review.opendev.org/74977915:40
openstackgerritMerged openstack/ironic-lib stable/victoria: Update TOX_CONSTRAINTS_FILE for stable/victoria  https://review.opendev.org/74978015:40
*** gyee has joined #openstack-ironic15:44
cdearborndtantsur: aok - will try to finish it up - thx!15:45
*** Qianbiao has quit IRC15:46
TheJuliaafaik, we are not unless there was a consensus I missed16:03
TheJuliain feature freeze that is16:03
cdearbornTheJulia, thx.  Are you trying to cut a sushy release today?16:04
openstackgerritDmitry Tantsur proposed openstack/ironic-python-agent master: [WIP] Generate a TLS certificate and send it to ironic  https://review.opendev.org/74993016:05
dtantsurcdearborn: today or on Monday16:05
dtantsurthis the global openstack limitation (non-client library release deadline)16:05
cdearborndtantsur: ok - thx16:05
*** lucasagomes has quit IRC16:06
*** dtantsur is now known as dtantsur|afk16:06
dtantsur|afkhave a good weekend everyone16:06
cdearborndtantsur: you too!16:06
NobodyCamGood Morning Folks, Glade to be back on IRC oh and Of course:16:07
NobodyCamTGIF!16:07
cdearbornGood morning NobodyCam!  Can't wait for the weekend16:08
TheJuliagood morning NobodyCam16:13
TheJuliagoodnight dtantsur|afk16:13
NobodyCamGood Morning cdearborn and TheJulia16:14
TheJuliaclarkb: is there anything fundimentally different between the environmnet variables when passed in the pre and run phases?16:16
clarkbTheJulia: there shouldn't be. They are separate playbooks so if you are setting them within a playbook they don't carry over to the next16:16
clarkbbut the behavior of setting them within a playbook should be the same if you set them again in the next playbook16:16
TheJuliaThat is what I figured16:18
TheJuliathen I wonder if neutron broke metalsmith16:18
TheJuliawhich would be awesome timing16:18
*** bnemec is now known as beekneemech16:18
*** dougsz has quit IRC16:33
*** derekh has quit IRC16:34
openstackgerritMerged openstack/metalsmith master: Add Python3 wallaby unit tests  https://review.opendev.org/74993916:47
*** ijw has quit IRC16:51
*** ijw has joined #openstack-ironic16:51
*** ijw_ has joined #openstack-ironic16:54
*** ijw has quit IRC16:57
TheJuliasigh, there must be something different17:02
openstackgerritJulia Kreger proposed openstack/ironic master: Remove token-less agent support  https://review.opendev.org/74940717:09
*** ijw has joined #openstack-ironic17:16
*** ijw_ has quit IRC17:20
*** bdodd_ is now known as bdodd17:22
openstackgerritJulia Kreger proposed openstack/ironic-inspector master: WIP: Power off before inspection  https://review.opendev.org/74984518:03
*** Lucas_Gray has quit IRC18:10
*** zzzeek has quit IRC18:19
*** zzzeek has joined #openstack-ironic18:20
*** ijw has quit IRC18:35
rpiosoTheJulia: I have a couple of concerns about https://review.opendev.org/749236.18:52
patchbotpatch 749236 - sushy - Add BIOS update apply time and maintenance window (MERGED) - 3 patch sets18:52
rpiosoTheJulia:  1) It hasn't passed the Dell EMC sushy Third-Party CI.18:52
rpiosoTheJulia: 2) The co-authors list is incomplete. mraineri (Michael Raineri) should be listed, too. I am not certain the syntax of the two co-authors is correct, because they are missing angle brackets.18:52
TheJulia*sigh*18:53
TheJuliaSo bottom line I can revert it, but it won't be in Victoria. Full stop.18:53
rpiosoTheJulia: Yikes! Well, the first concern may not matter much, because the CI doesn't test BIOS. Could we revert and clean-up the co-authors so Mike gets the recognition/credit he deserves?18:54
TheJuliaPropose the patches and I'll approve them.18:55
openstackgerritRichard G. Pioso proposed openstack/sushy master: Revert "Add BIOS update apply time and maintenance window"  https://review.opendev.org/75001718:57
TheJulia(that way it is not just me doing it on my own)18:57
iurygregoryI'm still around so I can +218:57
iurygregory=)18:57
TheJuliaWoot18:58
TheJuliaI think I'm going to go to the store... Trying to fix an inspector bug is making my head hurt.18:58
iurygregoryI'm still looking at the 700 lines patch =)18:58
TheJuliaThat one is not bad really. I had one gripe but I couldn't find usage of it so *shrug*18:58
*** qthepirate has joined #openstack-ironic18:59
*** priteau has quit IRC19:00
zer0c00lI don't think the lookup API docs reflect the 'agent_token' information https://docs.openstack.org/api-ref/baremetal/index.html?expanded=agent-lookup-detail,agent-heartbeat-detail19:01
zer0c00lThe example response body of v1/lookup doesn't have agent_token19:02
TheJuliaWoot19:03
TheJuliaI knew I forgot something!19:03
zer0c00lI was looking up how to heartbeat with the anaconda driver, the code seems to be adding the token but the API docs doesn't reflect that19:04
zer0c00lTheJulia: do you want me to open a docs  bug?19:04
TheJuliaEh... I guess.19:05
TheJuliaMight be faster just to propose a docs patch :)19:05
TheJuliarpioso: you'll need to strip the commit Id out to reupload19:08
TheJuliaOnce they are stacked, I can approve them19:08
rpiosoTheJulia: Almost done ...19:08
TheJuliacdearborn: looks like there is an issue with your latest sushy patch upload :(19:08
cdearbornTheJulia: checking...19:09
bdoddcdearborn TheJulia Looks like a conflict between the BIOS patch and the f/w update monitoring patch19:19
cdearbornbdodd, TheJulia, y - rebasing...19:20
openstackgerritRichard G. Pioso proposed openstack/sushy master: Add BIOS update apply time and maintenance window  https://review.opendev.org/75002019:22
rpiosoTheJulia: Done19:23
rpiosoTheJulia: Thank you!19:23
rpiosocdearborn: You may want to rebase to https://review.opendev.org/75002019:24
patchbotpatch 750020 - sushy - Add BIOS update apply time and maintenance window - 1 patch set19:24
*** qthepirate has quit IRC19:30
openstackgerritChristopher Dearborn proposed openstack/sushy master: Allow monitoring progress of a firmware update  https://review.opendev.org/74595019:33
openstackgerritRichard G. Pioso proposed openstack/sushy master: Revert "Add BIOS update apply time and maintenance window"  https://review.opendev.org/75001719:35
openstackgerritRichard G. Pioso proposed openstack/sushy master: Add BIOS update apply time and maintenance window  https://review.opendev.org/75002019:39
TheJuliarpioso: approved19:44
rpiosoTheJulia: Thank you so much!19:45
TheJuliacdearborn: a straight rebase didn't do it19:52
TheJuliazuul status page shows all of the unit tests failing19:52
cdearbornTheJulia, i tried to rebase on top of 750020, but it failed with conflicts and helpfully didn't show me what any of the conflicts were19:53
cdearbornTheJulia, will give it another shot19:53
TheJuliacdearborn: I wonder if you run a fresh unit test run locally if they will pass or not for yourself19:53
cdearbornTheJulia, where do you see the failures? looks like the test haven't been run yet: https://review.opendev.org/#/c/745950/19:56
patchbotpatch 745950 - sushy - Allow monitoring progress of a firmware update - 18 patch sets19:56
bdoddcdearborn After the rebase, you need to re-add the import statement for res_maps (in settings.py)19:57
bdodd(At least I think that is the issue)19:58
rpiosobdodd: Please note the use of @Redfish.SettingsApplyTime in the BIOS changes, both sushy and redfish driver. I expect analogous use of @Redfish.OperationApplyTime will be needed to get your proposed RAID changes to work against iDRACs.20:03
openstackgerritChristopher Dearborn proposed openstack/sushy master: Allow monitoring progress of a firmware update  https://review.opendev.org/74595020:04
cdearbornbdodd, thanks!  I made that change and tox runs clean now20:04
bdoddrpioso Thanks for the heads-up. Will look at it.20:04
rpiosobdodd: You're most welcome! Thanks for assisting with the BIOS stuffs. Very helpful!20:05
cdearbornTheJulia, will hopefully be gooder this time around...20:06
bdoddcdearborn \o/20:06
*** zzzeek has quit IRC20:06
*** zzzeek has joined #openstack-ironic20:10
TheJuliacdearborn: you'll love this link: https://zuul.opendev.org/t/openstack/status20:10
TheJuliaif you click the box for sushy, change set 745950 you'll see unit tests are passing20:11
TheJuliaor, to be more precise now, have passed20:11
cdearbornTheJulia, nice!  definitely going to bookmark that one!!!20:12
rpiosoTheJulia: Will we be meeting on Monday (Labor Day in US)?20:13
TheJuliaThere should, but I think everyone is taking it off20:29
TheJuliaeven folks in the EU are taking monday off afaik20:29
openstackgerritJulia Kreger proposed openstack/ironic-inspector master: Power off before inspection  https://review.opendev.org/74984520:32
TheJuliarpioso: I guess I'll show up and if people want to have a meeting then we'll have it, if people don't show its not a big deal, we'll deffer the agenda for a week I guess20:33
TheJuliaerr20:33
TheJuliahmm20:33
TheJuliaOpenInfra summit cfps for forum is open :\20:33
rpiosoTheJulia: We could defer the subjects you and I discussed late-ish Wednesday for a week.20:34
TheJuliayeah20:36
* TheJulia goes off to go to the store since grumbling about the metalsmith ci is not going to fix it20:36
TheJuliaiurygregory: I've left a +2 on cdearborn's patch, +a if you are happy with it20:46
iurygregoryTheJulia, I'm happy was waitting for CI to +120:47
iurygregorysome weird errors before =)20:47
iurygregoryI will let a +2 (since the result for the job may take a while - but at least know the unit jobs are green)20:48
TheJuliayeah, I was thinking the same20:49
*** Lucas_Gray has joined #openstack-ironic21:00
*** Lucas_Gray has quit IRC21:08
*** Lucas_Gray has joined #openstack-ironic21:14
*** rh-jelabarre has quit IRC21:29
openstackgerritMerged openstack/sushy master: Revert "Add BIOS update apply time and maintenance window"  https://review.opendev.org/75001721:34
openstackgerritMerged openstack/sushy master: Add BIOS update apply time and maintenance window  https://review.opendev.org/75002021:34
*** rloo has quit IRC21:36
* cdearborn just got back from a beer run21:36
*** rcernin_ has joined #openstack-ironic21:52
*** hjensas has quit IRC22:10
cdearbornTheJulia, iurygregory: yay!  unit tests are good to go now on https://review.opendev.org/#/c/745950/22:26
patchbotpatch 745950 - sushy - Allow monitoring progress of a firmware update - 19 patch sets22:26
iurygregorycdearborn, +A22:27
iurygregoryfunctional tests are green =)22:27
iurygregoryTheJulia, since monday is holiday in the US do you want to cancel the weekly meeting?22:28
*** rcernin_ has quit IRC22:29
*** dking has quit IRC22:36
TheJuliaiurygregory: I'd honestly prefer to at least try and have it22:44
TheJuliaiurygregory: that way I can at least do annoucements.22:45
iurygregoryTheJulia, ++ =)22:51
*** tosky has quit IRC23:07
*** Lucas_Gray has quit IRC23:12
openstackgerritJulia Kreger proposed openstack/metalsmith master: Rework jobs so failed devstack does not retry the job  https://review.opendev.org/74975823:24
TheJuliacdearborn: rpioso: I took another look at https://review.opendev.org/#/c/748927 and worst comes to worst once CI is back up and it can be ran past it, we can backport it as a butfix I think23:25
patchbotpatch 748927 - sushy - Make message parsing more resilient - 3 patch sets23:25
TheJuliajust create a bug to cover the MessageKey difference23:26
openstackgerritMerged openstack/sushy master: Allow monitoring progress of a firmware update  https://review.opendev.org/74595023:28

Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!